面向城市地震风险评估的人口空间分布研究  

Study on population spatial distribution for urban seismic risk assessment

作　　者：朱玉群 肖东升[1,2] ZHU Yuqun;XIAO Dongsheng(School of Civil Engineering and Surveying and Mapping,Southwest Petroleum University,Chengdu 610500,China;Disaster Prevention and Emergency Research Center for Surveying,Mapping and Remote Sensing,Southwest Petroleum University,Chengdu 610500,China)

机构地区：[1]西南石油大学土木工程与测绘学院,成都610500 [2]西南石油大学测绘遥感地理信息防灾应急研究中心,成都610500

出　　处：《测绘科学》2023年第9期262-272,共11页Science of Surveying and Mapping

基　　金：四川省区域创新合作项目(23QYCX0053)。

摘　　要：针对现有人口空间分布研究成果大多是格网数据,无法反映建筑物内的人口数量的问题,提出了一种新的计算不同时段建筑物内人口数量的方法:(1)定位率算法融合年龄比重,算法改进前后的人口估算数据与人口普查数据回归分析得出决定系数分别为0.9368、0.9471;(2)基于1 km格网人口数据和POI核密度数据,采用主成分分析确定人口权重,利用泰森多边形对1 km格网数据重分配,得到100 m格网人口数据,街道尺度绝对误差明显降低;(3)基于100 m格网人口数据、在室率和建筑物的人口吸引力的时序值,实现了贴合实际的建筑物尺度的人口分布。结果表明:改进定位率算法提高了人口估算值精度;POI核密度数据有助于反映城市人口分布的内部结构;建筑物尺度人口分布算法去除了室外人口,反映不同时间建筑物内人口数量,为地震应急救援力量部署决策提供数据支持。A new method for calculating the population inside buildings at different time periods was proposed to address the issue that most of the existing research results on population spatial distribution were grid data that could not reflect the population inside buildings.Firstly,the localization rate algorithm integrated the age proportion,and the regression analysis of the population estimation data before and after the algorithm improvement and the census data showed that the determination coefficients are 0.9368 and 0.9471,respectively;Then,based on 1 km grid population data and POI core density data,principal component analysis was used to determine the population weight,and Tyson polygon was used to redistribute 1 km grid data to obtain 100 m grid population data,and the absolute error of street scale was significantly reduced;Finally,based on the 100 m grid population data,the in-building rate and population attractiveness of buildings,the building scale population distribution that fitted the actual was realized.The results showed that the improved localization rate algorithm had improved the accuracy of population estimation;POI kernel density data helped to reflect the internal structure of urban population distribution;The building scale population distribution algorithm removed outdoor population and reflected the number of people inside the building at different times,providing data support for the deployment decision of earthquake emergency rescue forces.

关 键 词：城市地震风险评估 建筑物 POI 主成分分析 在室率 

分 类 号：P208[天文地球—地图制图学与地理信息工程]